Coverage-Adaptive Random Walks for Fast Sensory Data Collection

Random walks in wireless sensor networks can serve as fully local, very simple strategies for sink motion that reduce energy dissipation a lot but increase the latency of data collection. To achieve satisfactory energy-latency trade-offs the sink walks can be made adaptive, depending on network parameters such as density and/or history of past visits in each network region; but this increases the memory requirements. Towards better balances of memory/performance, we propose two new random walks: the Random Walk with Inertia and the Explore-andGo Random Walk; we also introduce a new metric (Proximity Variation) that captures the different way each walk gets close to the network nodes. We implement the new walks and experimentally compare them to known ones. The simulation findings demonstrate that the new walk's performance (cover time) gets close to the one of the (much stronger) biased walk, while in some other respects (partial cover time, proximity variation) they even outperfo...

• Type: Conference
• Year: 2010
• Where: ADHOCNOW
• Authors: Constantinos Marios Angelopoulos, Sotiris E. Nikoletseas, Dimitra Patroumpa, José D. P. Rolim
• Keywords: ADHOCNOW 2010, Computer Networks, Proximity Variation, Random Walks, Wireless Sensor Networks